标签:pen 日志 情况 添加 alt lis 实施 state key
--测试环境:
OS:Redhat linux(64)
Primary:
IP:192.168.94.198
SID:dgdb1
Hostname:dg1
DB_UNIQUE_NAME:dgdb1
Database:10.2.0.1(64)
Standby:
IP:192.168.94.199
SID:dgdb1
Hostname:dg2
DB_UNIQUE_NAME:dgdb1_s
Database:10.2.0.1(64)
--实施DG前的准备工作
--开启数据库logging及数据库archivelog
--开启数据库logging
SQL> alter database force logging;
--检查数据库是否开启archivelog
SQL> archive log list;
--如果数据库未开启archivelog,则要开启archivelog
SQL> shutdown immediate;
SQL> startup mount;
SQL> alter database archivelog;
--检查数据库是否开启archivelog
SQL> archive log list;
--创建相应目录(根据具体情况,primary和standby端要一致)
--Standby:
mkdir –p /export/home/oracle/product/10.2.0/oradata/dgdb1
mkdir –p /export/home/oracle/product/10.2.0/admin/dgdb1/adump
mkdir –p /export/home/oracle/product/10.2.0/admin/dgdb1/bdump
mkdir –p /export/home/oracle/product/10.2.0/admin/dgdb1/cdump
mkdir –p /export/home/oracle/product/10.2.0/admin/dgdb1/udump
mkdir –p /export/home/oracle/product/10.2.0/admin/dgdb1/dpdump
mkdir –p /export/home/oracle/product/10.2.0/admin/dgdb1/pfile
mkdir –p /export/home/oracle/archive
mkdir -p /export/home/oracle/bak
--primary:
mkdir –p /export/home/oracle/archive
mkdir -p /export/home/oracle/bak
--修改或新增listener.ora 和tnsnames.ora(或者用GUI工具配置)
--注意:listener.ora 中新增的部分在括号内而非括号外
--primary端:
--listener.ora:
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(SID_NAME = PLSExtProc)
(ORACLE_HOME = /export/home/oracle/product/10.2.0)
(PROGRAM = extproc)
)
(SID_DESC =
(GLOBAL_DBNAME = dgdb1)
(ORACLE_HOME = /export/home/oracle/product/10.2.0)
(SID_NAME = dgdb1)
)
)
LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = dg1)(PORT = 1521))
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC0))
)
)
--tnsnames.ora
dgdb1 =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.94.198)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = dgdb1)
)
)
dgdb1_s =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.94.199)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = dgdb1)
)
)
--standby端:
--listener.ora:
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(SID_NAME = PLSExtProc)
(ORACLE_HOME = /export/home/oracle/product/10.2.0)
(PROGRAM = extproc)
)
(SID_DESC =
(GLOBAL_DBNAME = dgdb1)
(ORACLE_HOME = /export/home/oracle/product/10.2.0)
(SID_NAME = dgdb1)
)
)
LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = dg2)(PORT = 1521))
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC0))
)
)
--tnsnames.ora
dgdb1 =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.94.198)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = dgdb1)
)
)
dgdb1_s =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.94.199)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = dgdb1)
)
)
--具体实施DG的步骤:
--在primary上生成pfile,并修改添加相应参数,生成standby需要的pfile
--primary端:
oracle$>sqlplus / as sysdba
SQL> CREATE PFILE=‘/export/home/oracle/standby.ora‘ FROM SPFILE;
--编辑生成的pfile文件(‘/export/home/oracle/standby.ora‘),添加如下参数:
*.db_unique_name=‘dgdb_s‘
*.fal_server=‘dgdb1‘
*.fal_client=‘dgdb_s‘
*.LOG_ARCHIVE_CONFIG=‘DG_CONFIG=(dgdb1,dgdb_s)‘
*.log_archive_dest_1=‘LOCATION=/export/home/oracle/archive VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=dgdb_s‘
*.LOG_ARCHIVE_DEST_2=‘SERVICE=dgdb1 LGWR ASYNC=40960 VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=dgdb1‘
*.LOG_ARCHIVE_DEST_STATE_1=‘ENABLE‘
*.LOG_ARCHIVE_DEST_STATE_2=‘ENABLE‘
*.standby_archive_dest=‘/export/home/oracle/archive‘
*.standby_file_management=‘AUTO‘
--修改完后拷贝到standby端
Oracle$Scp /export/home/oracle/standby.ora oracle@192.168.94.199:/export/home/oracle/
--在standby端使用pfile启动实例
--在primary端用命令创建数据库密码文件,并接拷贝至standby端相同路径下(如已存在,可直接拷贝)
--primary端手工创建数据库密码文件
Oracle$orapwd file=... password=...
--primary端拷贝数据库密码文件至standby端
Oracle$Scp /export/home/oracle/product/10.2.0/database/PWDdgdb1.ora oracle@192.168.94.199:/export/home/oracle/product/10.2.0/database
--standby端:
Oracle$set oracle_sid=dgdb1
Oracle$sqlplus / as sysdba
SQL> startup nomount pfile=’/export/home/oracle/standby.ora’
SQL> CREATE SPFILE FROM PFILE=‘/export/home/oracle/standby.ora‘;
--在primary端修改相应参数
--primary端:
Oracle$sqlplus / as sysdba
SQL> ALTER SYSTEM SET STANDBY_FILE_MANAGEMENT = AUTO scope=both;
SQL> ALTER SYSTEM SET fal_server=‘dgdb1_s‘ scope=both;
SQL> ALTER SYSTEM SET fal_client=‘dgdb1‘ scope=both;
SQL> ALTER SYSTEM SET LOG_ARCHIVE_CONFIG=‘DG_CONFIG=(dgdb1_s,dgdb1)‘ scope=both;
SQL> ALTER SYSTEM SET log_archive_dest_1=‘LOCATION=/export/home/oracle/archive VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=dgdb1‘ scope=both;
SQL> ALTER SYSTEM SET LOG_ARCHIVE_DEST_2=‘SERVICE=dgdb1_s LGWR ASYNC=40960 VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=dgdb1_s‘ scope=both;
SQL> ALTER SYSTEM SET LOG_ARCHIVE_DEST_STATE_1=‘ENABLE‘ scope=both;
SQL> ALTER SYSTEM SET LOG_ARCHIVE_DEST_STATE_2=‘ENABLE‘ scope=both;
SQL> ALTER SYSTEM SET standby_archive_dest=‘/export/home/oracle/archive‘ scope=both;
--在primary端开始使用rman备份数据库
--primary端:
Oracle$rman target /
RMAN>backup full format=‘/u01/app/oracle/bak/ora10g_%d_%T_%s‘ database include current controlfile for standby plus archivelog format=‘/u01/app/oracle/bak/arch_%d_%T_%s‘;
--备份完毕后,把相应备份文件拷贝到丛库的相应目录(目录必须一致)
Oracle$Scp /export/home/oracle/bak/* oracle@192.168.94.199:/export/home/oracle/bak/
--使用duplicate 还原standby数据库
--primary端:
Oracle$rman target / auxiliary sys/system@dgdb1_s
RMAN> duplicate target database for standby nofilenamecheck dorecover;
--收尾工作及开启dg
--standby端建立standby log(至少比redo多一组):
Oracle$sqlplus / as sysdba
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 4 (‘/export/home/oracle/product/10.2.0/oradata/dgdb1/redo04.log‘) size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 5 (‘/export/home/oracle/product/10.2.0/oradata/dgdb1/redo05.log‘) size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 6 (‘/export/home/oracle/product/10.2.0/oradata/dgdb1/redo06.log‘) size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 7 (‘/export/home/oracle/product/10.2.0/oradata/dgdb1/redo07.log‘) size 50M;
--开启服务(standby端):
SQL> alter database recover managed standby database disconnect from session;
--测试服务是否正常
--standby端(看归档日志号):
SQL> SELECT SEQUENCE#, FIRST_TIME, NEXT_TIME FROM V$ARCHIVED_LOG ORDER BY SEQUENCE#;
--primary端(强制一个日志切换):
SQL> alter system switch logfile;
--standby端(看归档日志号):
SQL> SELECT SEQUENCE#, FIRST_TIME, NEXT_TIME FROM V$ARCHIVED_LOG ORDER BY SEQUENCE#;
--standby端(看应用归档日志的号):
SQL> SELECT SEQUENCE#,APPLIED FROM V$ARCHIVED_LOG ORDER BY SEQUENCE#;
--实现primary、standby的切换:
--primary端:
SQL> SELECT SWITCHOVER_STATUS FROM V$DATABASE;
--primary端(上一步的结果必须是”TO STANDBY”才可以)
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PHYSICAL STANDBY;
--OR(上一步结果为”SESSIONS ACTIVE”,且解决不掉)
SQL>ALTER DATABASE COMMIT TO SWITCHOVER TO PHYSICAL STANDBY WITH SESSION SHUTDOWN;
--Primary端:
SQL> SHUTDOWN IMMEDIATE;
SQL> STARTUP MOUNT;
--standby端:
SQL> SELECT SWITCHOVER_STATUS FROM V$DATABASE;
--standby端(上一步的结果必须是”TO STANDBY”才可以):
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY;
--OR(上一步结果为”SESSIONS ACTIVE”,且解决不掉)
SQL>ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY WITH SESSION SHUTDOWN;
--打开新的primary:
SQL>ALTER DATABASE OPEN;
--OR(最后一次启动后以read only模式打开过)
SQL> SHUTDOWN IMMEDIATE;
SQL> STARTUP;
--在新standby上重新启动log apply services(如果必要):
SQL>ALTER DATABASE RECOVER MANAGED STANDBY DATABASE;
--OR(后台模式 )
SQL>ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT;
--OR(实时应用redo)
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE;
--新primary端(开始发送redo 数据到新standby端):
SQL> ALTER SYSTEM SWITCH LOGFILE;
标签:pen 日志 情况 添加 alt lis 实施 state key
原文地址:https://www.cnblogs.com/lhdz_bj/p/8759078.html